The Steno-2 study comes of age

In 1993, 160 people with type 2 diabetes and microalbuminuria were randomised in Denmark to an intensive or conventional treatment protocol. Those in the intensive arm attended the Steno specialist hospital every three months. As well as being provided with lifestyle advice on diet, exercise and smoking cessation, they all received renin–angiotensin system blockade therapy, a low dose of aspirin (after 5 years), vitamins A and E, folic acid and trace element supplements. They were also set more stringent blood glucose, blood pressure and blood lipid targets. At the end of the initial study period (median 7.8 years), there were significant reductions in microvascular and macrovascular endpoints in the intensive treatment group. All of the surviving patients were then set the same targets, but the intensive lifestyle interventions and supplements were stopped. By the next study point at 13.3 years, levels of glycaemia, blood pressure and lipids in the two groups were the same. What had happened to the participants after 21 years? A new study (summarised alongside) provides the answers.

In the two groups, 38 vs 55 participants had died (intensive vs conventional), with an absolute risk reduction (ARR) for mortality of 21%. The median survival benefit for intensive therapy was 7.9 years. For a first cardiovascular (CV) event, the ARR was 20% and median time was delayed by 8.1 years. In the intensive group 28 participants had no recorded CV event, compared to 13 in the conventional group. Looking at the Kaplan–Meier curves, there is a separation for CV events (including mortality) at around 4 years, and for mortality alone at 9–10 years. Cardiovascular event rates after the initial study period were more or less parallel, but mortality rates continued to diverge. There was a marked reduction in stroke and myocardial infarction (MI) rates in the intensive group. How were such remarkable results achieved with such a relatively small cohort?

In the intervention period there were modest reductions in glycaemia (absolute difference in HbA_1c, 8 mmol/mol; 0.7%), with greater reductions in systolic blood pressure (SBP) of 11 mmHg and serum low-density lipoprotein cholesterol (LDL-C) of 1.1 mmol/L. Disappointingly, there was no significant impact on smoking rates, BMI or lifestyle. By the time of the first post-intervention follow-up at 13.3 years, there was no longer any difference between groups, although we do not know at what precise time this occurred. If we assume that this happened midway, then the intervention group could have had more than 10 years of better glycaemia, blood pressure and lipid control. Using the UKPDS data (Stevens et al, 2004), the lower HbA_1c would have reduced the odds for an MI by around 12% and stroke by 26%. The SBP reduction would have reduced relative risk (RR) for a major CV event (including mortality) by ~13% (Ettehad et al, 2015) and the LDL-C reduction by 23% (Cholesterol Treatment Trialists’ [CTT] Collaboration et al, 2015). The lower SBP reduces the RR for stroke by nearly 29%. Most interesting is the finding that the difference between treatment groups was only seen in those with no history of a CV event over 21 years. Once an individual had experienced one, then there was no benefit of intensive therapy.

What can we take from all this? Firstly, the report highlights the high CV risk in these patients – 50% in the conventional group had a major CV event by 8 years and nearly 70% had died by 20 years. Secondly, improving glycaemia and lowering SBP and serum LDL-C all have a major beneficial effect on CV events in people with type 2 diabetes who are at relatively high risk. The data thus support current guidance with regard to treatment targets. Thirdly, primary prevention of a CV event seems critical. Finally, and perhaps most depressingly in the light of the current NHS Diabetes Prevention Programme, even in the context of a clinical trial delivered by an international centre of excellence, it is hard to influence lifestyle and smoking habits.

To read the article summaries, please download the PDF